[G] Map Pool Creation Guide

So here you are, you are a tournament admin and you have a tournament all set up. However, maps have not been set up. What do you do? Easy, follow this guide and in a few simple steps you will have a map pool that not only is very fun to play on, but also will be well received by both the community, players, and mapmakers alike.


No. There is quite a few issues with a map pool like this, and we will cover them below. Just think, as a organizer of a tournament you are putting your heart and soul into to make the best possible why spend so little time on something so major? This can influence the style of games of you tournament, the overall run time, and a million other things. Would you just randomly grab a tournament structure and format? Obviously not, and a good map pool is just as important if not more important then a format.

Let's start at square one, which maps are available to pick from. This is where you have the most choices in how you want to shape your map pool. The main decision you have to make comes down to do you make a traditional map pool or a innovative one. This will set the tone of your map pool and mindset going into the full creation stage, and is very important very much like having a clean opener in SC2 is to a player. Also like an opener, they have ups and down associated with them but offer different risk/reward factors.

Traditional Map Pool: This is the type that most major tournaments choose to go with modern day. This contains all maps that have been used in major tournaments or events before, and refuses to take chances with new maps.

Pros: Safe picks that are tried and tested, predictable game times, Non-Korean players will enjoy not having to learn new maps.
Cons: Expect your tournaments games to look like every other tournament, fans and mapmakers will complain.

Example Traditional Map Pool:
+ Show Spoiler +

Innovative Map Pool
This is the true rarity in the Starcraft II scene. It focuses on using some of the traditional maps, but cycles out old maps in favor of new unexplored maps. While this contains more risks, it often contains even bigger rewards is great games. As far as I know only myself with the Korean Weekly and the GSL are using this style. Other orgs recently have been trying these types of pools, but we have not seen if they will continue to do so, or they did not continue to do so.

Pros: Contains some tried and tested maps while pressing boundaries, fans and mapmakers will applaud you.
Cons: Non-Korean players will complain about having to learn new maps or not train on them properly, more variance in game time.

Example Innovative Map Pool:
+ Show Spoiler +

Ok so you have seen both types, now you start your decision here. Do I want to stay standard and play it safe or do I want to take a risk with possibly a great reward? Neither choice is a wrong choice.

One of the longest and biggest debates, "how many maps should I pick to compose my pool?" There is also no right answer to this one yet, however there is wrong answers. I will cover this section by going over each size and what is good/bad about it.

The one thing to keep in mind and that I will be showing here is how much practiced is required with each size. All players have different training regiments, but until you have played a map 20+ times (except for Gumiho or TaeJa), you will not be playing the map optimally. After 20 games, you may still not be using 100% of the map 100% right, but you will know a loy of the layouts, trick spots, and a lot of other map specific info.

8+ Maps:



Want to screw up your map pool? Put in this many maps. There is not a single excuse in the world for having a map pool this big. You have to add in artificial rules like double veto's just to make it work. Worse is that NO ONE is going to practice the entire pool, they are going to only practice a portion of it properly. If you are thinking of making a map pool of this size, stop right here and shrink it. If you are an organization that has both a individual league and a team league, have separate pools like the GSTL. Putting the responsibility of learning this many maps on a single player is too much, even for tournaments with massive prize pools. You will get a lower quality of game, and have to add in a slew of rules to accommodate for this. Any tournament that has this size of pool should immediately look to shrink it. If you have a 11 map pool, you need to scratch the entire pool, and start from square one as any modifications to bring it down to 7 or below will unbalance the style and scale of the pool.

7 Maps:
The golden standard of Starcraft II. For the longest time I personally believed this was the best way to do a pool. Allows for a BO7 with no repeats of maps. Still a big map pool, but not too big. If you choose a seven map pool, you will do alright. It will offer lots of variety in picks, and will likely not contain any racial imbalances that cannot be fixed with vetos.

However one thing I came to realize over time is that almost every tournament only has one BO7 series in the entire event. The finals. What I mean is that the best reason and the reason seven map pools became the golden standard is you potentially have a bo7 on all unique maps, no repeats.

This concept was designed when the map pools contained maps (Steppes of War, Kulas Ravine, close positions maps) that very heavily favored certain races (Terran) for a long period of time when no one knew how to play the game right, let alone abuse the maps (CatZ excluded). So players developed ways to play against a blackout Terran map pool, and individual players got good on different maps with different strategies. Now the overall map pool is a lot more fair and balanced, and some maps favor one race/match up (Sanshorn Mist T>Z), and some maps seem to have pretty even stats and just fluctuate with the metagame (Daybreak). There is more stability, and just including Cloud Kingdom, Daybreak, Entombed Valley, and Antiga Shipyard 1.2 (no golds, cross only) can provide a fair and balanced experience by themselves. By introducing three maps outside this core four on top, you are introducing more instability into your own tournament for no payoff.

No matter what tournament players are playing in, they practice those four maps (and historically there has always been a core four) leaving more time to prepare for your tournament. Instead of them being able to focus on the one non-core map they now need to focus on three, leading to sub-par preparation normally. You are rewarding the better player while still introducing a mechanic of that player having to go outside the norm to practice for your event.

If you already have a seven map pool, that's good however you should look to "trim some of the fat" with a 5 map pool as described below, but do not worry too much, a seven map pool is still very good.

5 Maps:
This is a new size of map pool that has not traditionally used in Starcraft, but been theorized on by notable community members on State of the Game like Artosis and Liquid`Nony. This is the size of map pool I have been using for my tournament, the ESV TV Korean Weekly for about 6 months now. I personally also believe this is the ideal map pool size for all tournaments, big or small in Starcraft II. Unlike in BW, BO7 is something that is actively used on a professional level and a four map pool would end with a lot of repeats. With a 5 map pool you only have a max of two repeats per five maps. In the event of a BO5, there is no repeat and all five maps are used.

Creating a smaller pool allows you to properly introduce new maps. The players will run into them more often in the tournament, they will have less to practice (since every pool has three+ known maps, even mine), and will get more time to figure out said map.

I think if you could teleport in time and see what kind of map pools are being used five years from now, this will be the type. It is easy for fans to follow, easy for players to practice, and easy for tournament's to introduce new maps to make themselves different from the rest. If in the future the BO5 standard for finals takes over however, this may all change. I think all tournaments should be looking as soon as possible to move to a five map format. If four maps ends up being the future, the change from five to four will be much smoother compared to the one from seven to four.

Relevant video


4 Maps:
Many people are probably wondering, why 4 maps? No one uses it? Well KeSPA does and in their first season in SC2 they are sticking with what they did in BW with SC2 at this point. So it needs to be addressed This is a very interesting format, and worked VERY well for the BO5 heavy BW leagues. It very well stand to be the better option for SC2, and until the proleague starts we really won't know, and even then it will take several seasons to be sure.

I admit I have never used a a map pool of only four maps, however I still feel that it would be an acceptable standard for tournaments.

3 Maps:
This happens from time to time with GSL qualifier related events. There is not much to talk about, if you intend to broadcast the tournament ever in any way shape or form, don't do this. If it's not meant to be broadcasted and will contain a large number of amateur players it can work. I cannot stress the point enough that your tournament loses most of it's appeal with these pools, and again should NEVER be used in anything that is broadcasted.

These pools also also allow for a bad pick of maps to influence a tournament deeply. Pick two maps that are heavily Terran favored, and you should end up with Terrans taking most of the Top 8, same with other races. You want racial diversity as much as possible and a three map pool decreases that.]

This is where the fun begins.

2 player/4 player split
This should be one of the first concerns, and this is how you will help control game length so you don't have a tournament you were expecting to run in five hours take 15.

This is a pretty easy part really. The more four player maps you have, the longer the tournament will take. The reasoning being of course that on four player maps proxy'ing is even more risky than normal and it's only a move a player tends to use when up multiple games in a series, or when matched against an opponent they do not think they can beat in a straight up game.

You should always have at least two or more two player maps. Unless you have a very small tournament player wise (IEM for example) you should have more two player maps then four player. This will allow the tournament to flow better, and prevent all the two player maps from being veto'd (unless you are playing 8+ maps with double vetoes, sigh). You should never have all two player or all four player maps.

Map Size
This is something that gets very overlooked. At ESV we personally classify maps in three main categories:
1: Rush Maps
2: Medium Maps
3: Macro Maps
You should be looking to fill all the categories with a minimum of two "rush maps". This allows cheese to be a part of your tournament, and it is an important part of a tournament. This also forces your players to be diverse, and not be able to rely on only hitting large maps and winning off superior macro. The remaining 1-3 slots (depending on size of your map pool) should be depending on your tournament size and timeframe.

Examples:
Rush Maps: Ohana, Korhal Compound
Medium Maps: Cloud Kingdom, Daybreak
Macro Maps: Metropolis, Entombed Valley

Introducing new maps
I will make this simple because I want this to stick.

You will NEVER ruin a tournament by introducing 1-2 new maps (unless your map pool is three maps).

NEVER introduce more then two new maps in a tournament map pool. The ONLY exception is when updating versions (ie: Korhal Compound LE >>>> Korhal Compound TE) of commonly used maps.

Map age and quality of game
At a certain point in a maps life the map reaches the end of it's life. When this happens the map still "works" per se. You get games that play like standard games and they are ok. On a map that is not outdated, you will get a large number of "great" or "epic" games. On on outdated map, you will get many less. Every tournament wants to have those memorable games that everyone remembers, you are decreasing your odds of doing so by keeping in old maps. Many games will blur together with the 3000+ other games the viewer has seen on that map.

You are putting most of the time players that play this game for eight or more hours a day that do this for a profession, they can make a spectacular game on any map, but you can make more on a fresh map pool. Just because one game was great (for example Stephano/Kiwikaki on Shattered Temple) does not mean the map is great. It means Stephano and Kiwikaki is a great stylistic match up between two of the most creative minds of their respective races in the most violate non-mirror match up. They would have put on a show on Steppes of War, but a good map that does not make it.

When does a maps life start and end? A start point would be when it was accepted into ladder or major tournaments for the first time. Once a map has reached 8-12 months from that point, it should be looked at "Is the map producing interesting and fresh games?" needs to be the question you ask.

Now to call out a few, in my opinion, none of these maps should ever be used in a tournament that involved any type of a prize ever again because they have stagnated and the average quality of game has gone gone a lot. If you are a tourney admin and have one of these maps, you should remove it and look for a suitable replacement right away. You will only be benefiting your players, fans, and yourselves by doing so.

Shattered Temple 
Sanshorn Mist AE
Shakuras Pleateau
Terminus
Metalopolis
Bel'Shir Beach
Crossfire
Tal'Darim Altar
Dual sight
Xel'Naga Caverns
Testbug
Xel'Naga Fortress
Crevasse

Which version to use?
Ask the mapmaker! They will not bullshit you. If the LE is better, they will tell you. If the GSL version is the best they will tell you. These maps represent them, and they want them to play the best it can. When in doubt, don't use the LE.

As some general guidelines however you should look to use the version the map creator is actively using/pushing. An example of this would be Metropolis. Several revisions have been made to it. First was the stuff to fix the lag issues, then the removal of the island to prevent 0 supply wins. However there is a version with fixes for both of these, and it's one of the original versions! 대도시 Lite is a version of the map that has been published for a long time on the Korean server. It contains some fixes for the lag issues, and also has neutral tumors on the islands that require a unit with an attack and detection to kill. Many problems could have been solved by quickly asking the mapmaker about these issues and the fixes, and finding out everything is fixed in an alternate version.

Another example is tournaments that use Korhal Compound instead of the tournament focused Korhal Compound TE. Mapmakers hate having multiple versions of the same map generally and if they feel that a TE is worth it, there is often a reason behind it. Check with the mapmaker, have them explain why they made these changes or what your concerns are, most of the time the smallest things may be something that an entire map is designed around, and you can destroy the main idea behind a map without even realizing or meaning to if you just go edit the map on your own. Let the mapmakers work on your concerns with you, instead of grabbing whomever spent 12 minutes in the editor that one time in the office and having him edit it. Having 43 versions of one map is bad for everyone.